I have been asked to set up multiple database replication which I have
done before for simple cases however there are some nuances with this
instance that add some complexity and I'd like to hear your collective
expertise on this proposed scenario:-
1) Single master database
2) n (probably 3 to start with) number of slave databases
3) All but 5 tables (123 tables in total) are to be replicated from the
master to all the slaves
4) 3 tables from the slaves are to be replicated back to the master
It is mainly item 4) that concerns me - the primary ID's are almost
certain to collide unless I seed the auto increment ID to partition the
IDs into separate ranges or does MySQL handle this issue?
There are some foreign keys on one of the 3 slave to master tables but
they are pointing at some extremely static tables that are very unlikely
Is the above a feasible implementation...?
Thanks in advance for any advice/pointers!